PLE with off-line clean-up and SPLE of theCRMgave RSDs
about 10–15%. Additionally all measurements obtained for each
analyte by SPLE fall within the prediction interval given in the
certificate of analysis. This demonstrates the good precision and
accuracy of the SPLE method. Furthermore, the SPLE results
have confidence intervals that include or overlap with a wide
range of the confidence intervals of theCRMexcept for -HCH,
-HCH and p,p-DDE which are somewhat above the specified
interval. In contrast to this, -endosulfan was observed to be
the only analyte having a yield far below the confidence interval
of the CRM. In fact the same observation was noted and
reported in earlier works [11,19]. The three different techniques
of extraction employed in this experiment (sonication, PLE with
off-line clean-up and SPLE) uniformly gave lower recovery
for -endosulfan. Thermal effects on -endosulfan from PLE
and GC conditions have already been ruled out [19,27]. Therefore,
reasonable justification for the lower recovery observed
may not be inadequate analytical methods of extraction rather
it could possibly be due to the conversion of -endosulfan into
the major metabolite through time within the certified reference
soil sample itself [19].